Tobacco-specific nitrosamines (TSNAs) exist in environmental waters; however, it is unknown whether TSNAs can be produced during water disinfection. Here we report on the investigation and evidence of TSNAs as a new class of disinfection byproducts (DBPs). Using five common TSNAs, including (methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) as the targets, we first developed a solid phase extraction (SPE) and liquid chromatography-tandem mass spectrometry (LC-MS/MS) method capable of rapidly determining these TSNAs at levels as low as 0.02 ng/L in treated water. Using this highly sensitive method, we investigated the occurrence and formation potential (FP) (precursor test conducted in the presence of chloramines) of TSNAs in treated water from two wastewater treatment plants (WWTPs) and seven drinking water treatment plants (DWTPs). NNAL was detected in the FP samples, but not in the samples before the FP test, confirming NNAL as a DBP. NNK was detected in the treated wastewater before the FP test, but its concentration increased significantly after chloramination in two of three tests. Thus, NNK could be a DBP and/or a contaminant in wastewater. Moreover, these TSNAs were detected in FP tests of wastewater-impacted DWTP plant influents in 9 of 11 samples. However, TSNAs were not detected at full-scale DWTPs, except for at one DWTP with high ammonia where breakpoint chlorination was not achieved. The concentration of the sum of five TSNAs (0.3 ng/L) was 100-fold lower than NDMA, suggesting that TSNAs have a minor contribution to total nitrosamines in water. We examined several factors in the treatment process and found that chlorine or ozone may destroy TSNA precursors and granular activated carbon (GAC) treatment may remove the precursors. Further research is warranted into the efficiency of these processes at different DWTPs using sources of varying water quality.